Plasma Cells: Finding New Light at the End of B Cell Development Kathryn L

Home , B cell, Germinal center, Lymphopoiesis, Naive B cell, Plasma cell

Plasma cells: finding new light at the end of B cell development Kathryn L. Calame

Plasma cells are cellular factories devoted entire- Upon plasma cell differentiation, there is a marked increase in ly to the manufacture and export of a single prod- steady-state amounts of Ig heavy and light chain mRNA and, when 2 uct: soluble immunoglobulin (Ig). As the final required for IgM and IgA secretion, J chain mRNA . Whether the increase in Ig mRNA is due to increased transcription, increased mediators of a humoral response, plasma cells mRNA stability or, as seems likely, both mechanisms, remains con- play a critical role in adaptive immunity.Although troversial2. There is also an increase in secreted versus membrane intense effort has been devoted to studying the forms of heavy chain mRNA, as determined by differential use of poly(A) sites that may involve the availability of one component of regulation and requirements for early B cell the polyadenylation machinery, cleavage-stimulation factor Cst-643. development, little information has been avail- To accommodate translation and secretion of the abundant Ig able on plasma cells. However, more recent mRNAs, plasma cells have an increased cytoplasmic to nuclear ratio work—including studies on genetically altered and prominent amounts of rough endoplasmic reticulum and secreto- ry vacuoles. mice and data from microarray analyses—has Numerous B cell–specific surface proteins are down-regulated begun to identify the regulatory cascades that upon plasma cell differentiation, including major histocompatibility initiate and maintain the plasma cell phenotype. complex (MHC) class II, B220, CD19, CD21 and CD22. MHC class This review will summarize our current under- II, which is required for antigen presentation to T cells, is absent due to lack of class II transactivator (CIITA), a coactivator required for standing of the molecules that regulate commit- transcription of MHC class II, invariant chain and HLA-DM4. The ment to a plasma cell fate and those that medi- proteoglycan syndecan-1, which recognizes extracellular matrix and ate plasma cell function. growth factors, is induced on antibody-secreting B cells and is often used as a marker of plasma cells5. Among integrin family proteins, © http://immunol.nature.com Group 2001 Nature Publishing which mediate cell-matrix and cell-cell adhesion functions, VLA-4 Plasma cells are interesting for several reasons. On a basic level, they (very late antigen 4) is most predominant on plasma cells. Perhaps show the general mechanisms involved in the commitment of somat- most important for differential homing of plasma cells, the ic cells to a terminally differentiated, post-mitotic state. As final chemokine receptors CXCR5 and CCR7 are decreased, which mediators of the humoral immune response, they provide insights reduces responsiveness to the B and T cell zone chemokines into the regulation of humoral immunity that may be helpful for vac- CXCL13, CCL19 and CCL216. In contrast, expression of CXCR4, cine development. They may also aid our understanding of certain which recognizes CXCL12 present in splenic red pulp, lymph node human diseases. For example, information about normal plasma cells medullary cords and in bone marrow, remains high6. These changes may help us understand human multiple myeloma, a disease that mediate movement of plasma cells from the follicles to other loca- involves malignant plasma cells. In addition, soluble antibodies or tions, including the bone marrow. immune complexes are pathological in many autoimmune diseases, Several transcription factors required for earlier steps in B cell devel- including myasthenia gravis, Grave’s disease, lupus erythematosus opment are decreased or absent in plasma cells. These include B cell and rheumatoid arthritis. Therefore, knowledge of plasma cell biolo- lineage–specific activator (BSAP), encoded by PAX5, which is required gy may eventually provide a rational basis for designing treatments for commitment to the B cell lineage7 and for B cell lymphopoeisis in for these diseases. bone marrow and spleen8,9. CIITA is also decreased in plasma cells4. Other transcription factors expressed at earlier stages in B cell devel- Characteristics of plasma cells opment, such as early B cell factor (EBF), A-Myb and BCL-6 are also Because they are terminally differentiated, end-stage cells, plasma decreased in plasma cells. However, a few transcription factors do cells do not divide. However, “plasmablasts” in extrafollicular increase in plasma cells. B lymphocyte–induced maturation protein 1 regions or exiting germinal centers (GCs) do undergo cell division (Blimp-1) is induced upon plasma cell differentiation10 and expressed just before they become plasma cells1. The lifespan of nondividing in vivo in plasma cells and a subset of GC B cells with a partial plasma plasma cells varies from a few days to many months and, as detailed cell phenotype11. Interferon-regulatory factor 4 (IRF4), which associ- below, this appears to be determined in part by their developmental ates with the Ets family protein PU.112, is also highly expressed in the history and homing capacity as well as by the finite capacity of the Blimp-1+ subset of GC B cells and in plasma cells13. X box–binding spleen to support them1. Other characteristics of plasma cells are sum- protein 1 (XBP-1), although ubiquitously expressed, is induced upon marized below (Fig. 1). activation of B cells and increased in plasma cells14.

Departments of Microbiology and Biochemistry & Molecular Biophysics, Columbia University College of Physicians and Surgeons, New York, NY 10032, USA. Correspondence should be addressed to K. L. C. ([email protected]).

Commitment to a plasma cell fate the other hand, xid and Btk–/– mice, which show compromised BCR Naïve B cell immigrants from the bone marrow mature in a variety of signaling, have MZ B cells but reduced survival of follicular B cells27. B cell compartments. Outside the bone marrow, spleen is a major site However, alternative explanations are possible. For example, for maturation of conventional B2 cells, whereas B1 cells mature pri- decreased MZ B cells in Aiolos–/– mice could result from enhanced marily in peritoneal and pleural cavities15. In the spleen, most lym- negative selection by self-antigen. phocytes are located in white pulp nodules that are interspersed in the Tumor necrosis factor (TNF) family ligands and TNF receptors red pulp matrix16. In each white pulp nodule, blood is supplied from (TNFRs) play critical roles in determining splenic architecture and B a central arteriole surrounded by a periarteriolar sheath (PALS). T cell development16. Lymphotoxins are necessary for the establishment cells are found primarily in the PALS, whereas most B cells cluster of splenic T cell and B cell zones and TNF is involved in the forma- around follicular dendritic cells in follicles adjacent to the PALS. A tion of marginal sinuses and GCs16. In addition, two ligands—B lym- marginal zone (MZ)—which contains specialized dendritic cells and phocyte stimulator (BLyS) and APRIL (a proliferation-inducing lig- macrophages, memory and naïve B cells—surrounds the outside of and)—and three receptors—TACI (transmembrane activator and each nodule and acts as an antigen-filtering and scavenging area. CAML interactor) and BCMA (B cell maturation antigen), which rec- When naïve B cells leave the bone marrow and enter the spleen, a ognize both, and BAFF-R (B cell–activating factor receptor), which maturation process involving both positive and negative selection recognizes BLyS28—have important roles in the development of ensues and most cells die in the outer PALS. Of the surviving B cells, mature follicular and GC B cells and in Ig production29. BlyS is the majority pass through the T cell zone and enter primary follicles required for maturation of follicular B cells and TACI–/– mice have ele- where they continue to mature into naïve follicular B cells vated B cells but lack T cell–independent type 2 B cell responses30. (IgMintIgDhiCD21intCD23hi)17. A smaller fraction are found in the MZ Further studies on TNF-TNFR proteins are likely to identify addition- as naïve MZ B cells (IgMhiIgDloCD21hiCD23lo)18. Follicular B cells al roles for them in B lymphoiesis and refine our understanding of how recirculate to bone marrow and lymph nodes but MZ B cells do not they affect humoral immunity and plasmacytic differentiation. recirculate. Antigen, either in soluble form A later response to antigen occurs when or associated with dendritic cells, activates follicular B cells are recruited by T naïve B cells in various B cell compart- cell–dependent antigens into GCs in spleen ments and, ultimately, leads to terminal and lymph nodes. It is also possible, but not plasmacytic differentiation. However, plas- yet established, that some MZ or B1 cells macytic differentiation occurs at different may enter follicles and mature in GCs. The times and in different locations during an GC reaction has been intensively studied31, immune response (Fig. 2). but will not be described in detail here. MZ and B1 cells respond most rapidly to Iterative cycles of proliferation, somatic antigen, giving rise to the first wave of plas- hypermutation and apoptosis in the GC © http://immunol.nature.com Group 2001 Nature Publishing ma cells and the earliest antibody response result in the selection of B cells that make to antigen19. B1 cells are a self-renewing an antibody with high affinity for cognate Figure 1. Plasma cell. Two characteristic surface mole- subset with a restricted repertoire that pro- cules, Syndecan-1 and CXCR4, three key transcription fac- antigen. In addition, isotype switching vides a response to common bacterial anti- tors, Blimp-1, IRF4 and XBP-1, and increased endoplasmic occurs during the GC reaction. The GC gens and plays a role in autoimmunity. For reticulum (ER) are shown. reaction usually peaks ∼10–12 days after example, B1 cells provide T cell–indepen- immunization, giving rise to two types of B dent production of mucosal IgA in response to commensal bacteria20. cells: nonsecreting memory B cells that are Ig surface–positive and The lineage relationship between B1 and B2 cells remains controver- antibody-secreting plasmablasts. Plasmablasts leave the GC and sial and will not be reviewed here. However, the size of the B1 versus develop into terminally differentiated plasma cells that secrete high- B2 compartments appears to depend on expression and composition affinity antibodies. Some of these plasma cells home to the bone mar- of the B cell receptor (BCR)18,21. row where they receive survival signals from stromal cells via cell When naïve MZ B cells encounter antigen, they become activated adhesion molecules32. Bone marrow plasma cells continue to secrete and leave the MZ within 1–3 days. They proliferate and form foci of antibody for many months and are sometimes considered to be a part plasma cells in extrafollicular regions along the periphery of the of the secondary response, although they do not respond to secondary PALS. Antibodies made by plasma cells that arise from MZ B cells antigenic stimulation33. are usually IgM and are often of low affinity22. Most of these early What signals tell a GC B cell to exit the GC? Because somatic plasma cells survive only about 3 days, undergoing apoptosis in situ, mutation, which drives affinity maturation, is a random process, it possibly due to down-regulation of the anti-apoptotic Bcl-2 homolog seems unlikely that a given number of cell divisions determines when A123. However, additional factors affect survival in the spleen and a a B cell leaves the GC. More likely, signaling via high-affinity anti- portion of the plasma cells in extrafollicular regions, possibly those body causes an end to the iterative cycles of mutation and selection, associated with dendritic cells, survive longer than 3 days1. although there is little experimental evidence to support this idea. Because antigen stimulation drives the development of MZ B cells What determines memory or plasma cell fate? One study of affinity toward early plasma cells, the factors that determine MZ versus fol- maturation in antibody-secreting versus memory cells suggests that licular B cell fate are key to understanding plasma cell development. expression of high-affinity antibody may favor plasma cell rather than Although BCR signals are critical for survival of all naïve B cells24, memory cell development34. Another informative approach has been to recent data suggest that migration to the MZ versus follicles may be study B cell differentiation in defined culture systems. These cultures determined by the strength of BCR signals, with lower intensity sig- have shown that signals via CD40 favor a memory phenotype, where- nals favoring MZ B cells25. Supporting this idea, MZ B cells are as in the presence of interleukin 2 (IL-2) and IL-10 and the absence of absent in Aiolos–/– mice, which have enhanced BCR signaling25,26. On CD40 ligand (CD40L), terminal plasma cell differentiation occurs35.

Signals from another TNFR, OX40L, also favor a plasma cell fate36. complementation system was used to study the effect of XBP-1 defi- IL-10 also interrupts memory B cell expansion, whereas IL-4 directs ciency in T and B cells14. XBP-1–/– T cells appeared to be normal and B cells to a memory fate37,38. Although it has pleiotropic activities, IL- the number and distribution of B cells in the bone marrow and spleen 6 is required for differentiation of plasmablasts and antibody secre- of chimeric mice was also normal. However, the chimeric mice had tion39, in part due to induction of the cyclin-dependent kinase inhibitor severely reduced concentrations of serum Ig. Upon closer analysis, p1840. IL-6 is of particular interest because it is also required for when XBP-1–/– splenic B cells were activated in vitro, the activation myeloma cell growth and survival41. Finally, differentiation and sur- markers CD44 and CD69 were induced normally, cells proliferated vival of plasma cells depends on, and may be limited by, specific types normally and germ-line transcripts appeared normally. This all sug- of plasmablast-associated dendritic cells42. gested that the XBP-1–/– B cells were fully capable of responding to In response to secondary antigenic stimulation, rapid and massive activation. However, the XBP-1–/– splenocytes did not secrete anti- clonal expansion of memory B body and did not express cells occurs, generating eight- Syndecan-1 on their surface to tenfold more plasma cells after activation in vitro. In than a primary response does43. addition, when the chimeric The biology of memory B cells mice were immunized with T remains somewhat mysterious, cell–independent (TI) or T although it is known that they cell–dependent (TD) antigens are long-lived, even in the or exposed to polyoma virus, absence of antigen44, and do not secretion of serum Ig of all iso- secrete antibody43. Upon leav- types was severely reduced. ing the GC, memory B cells Although it is expressed recirculate or home to draining ubiquitously, XBP-1 is induced areas of lymph node and upon activation of splenic B spleen. A distinct population of cells and is present at high con- B220– memory cells has been centrations in plasma cells14. It identified that undergoes affini- is repressed by BSAP50, a regu- ty maturation in the splenic red lator that decreases during pulp, is maintained long-term plasma cell differentiation9,51. in spleen and bone marrow and Therefore, reversal of BSAP is thought to be a major compo- repression is likely to account, nent of the splenic memory cell at least in part, for the induc- © http://immunol.nature.com Group 2001 Nature Publishing compartment45. Upon antigen tion of XBP-1. In addition, encounter, memory cells are XBP-1 is induced in response biased to become plasma to IL-6 signaling in myeloma cells46. Although CD40 is cells52 and by activating tran- important for the development scription factor 6 (ATF-6), of memory B cells, it does not which is activated in response appear to be required for the to endoplasmic stress53. These development of plasma cells in signals may also be important a secondary response47. In addi- for XBP-1 induction during tion to formation of plasma plasma cell development. cells, some memory cells must No target genes for XBP-1 be maintained during the sec- Figure 2. Generation of plasma cells in spleen during a humoral response. have been identified that would Solid orange arrows denote maturation pathways for B cells; dashed orange arrow ondary response to antigen. denotes uncertainty regarding migration of B cells from MZ to follicles. Blue arrows explain its critical role in plas- Signals influencing the deci- denote three occasions when plasma cells develop: after antigen stimulation of MZ B ma cells. Indeed, the time(s) sion between memory mainte- cells; after the GC reaction; and after secondary antigenic stimulation of memory B cells. when XBP-1 is required during nance and plasmacytic differen- This figure is simplified and does not indicate T cells, dendritic cells or the various the sequence of plasmablast to tiation are not well understood cytokines that participate in these processes. plasma cell development is not at present. yet clear. Blimp-1, which has been called a “master regulator of terminal B Key transcription factors in differentiation cell differentiation”10, was identified in a subtractive screen as a gene XBP-1 is the only transcription factor that is uniquely required for that was induced upon IL-2 + IL-5–dependent differentiation of the plasma cell differentiation but is not, apparently, required for any BCL-1 murine lymphoma to a plasma cell state10. A key activity of earlier stage of B cell development14. XBP-1 was first discovered the protein, first observed in BCL-1 lymphoma cells10 and subse- based on its ability to bind an “X box” sequence in the HLA-DRα quently confirmed in primary splenocytes54,55, is its ability to drive promoter48. It is a basic-leucine zipper protein that activates tran- activated or GC-like B cells to become antibody-secreting cells with scription. XBP-1 is ubiquitously expressed and mice deficient in a plasma cell phenotype. Both human and murine Blimp-1 proteins XBP-1 die as embryos due to defective hepatocyte growth, which use five zinc-finger motifs for sequence-specific binding to DNA56. 49 leads to reduced hematopoiesis and anemia . However, an unexpect- Blimp-1 is a transcriptional repressor and a proline-rich region NH2- ed finding was made when the RAG (recombination-activating gene) terminal to the zinc-finger domains is required for association of the

protein with hGroucho57 and class I histone deacetylases58 and for the row and for isotype switching in GCs8. Down-regulation of BSAP is protein to be able to repress transcription. required for development of antibody-secreting cells63 (K.-I. Lin et Blimp-1 is present in all plasma cells, including those formed in a al., unpublished data), probably because it represses XBP-150, J primary response to either TI or TD antigen, those formed from chain64 and Ig heavy chain gene transcription65. memory cells in a secondary response and in long-lived plasma cells However, repression of c-Myc, MHC2TA and Pax5 is unlikely to in human bone marrow11. Blimp-1 was not found in memory B cells explain the entire program of plasma cell development triggered by but is expressed in a subset (5–15%) of GC B cells. Cells in this sub- Blimp-166–68. With the use of cDNA microarray technology, a more set have a partial plasma cell phenotype because they lack expression complete understanding of the program of changes in gene expression of BCL-6 and BSAP, but express IRF4 and Sydecan-1, in addition to triggered by expression of Blimp-1 in human B cell lines has been Blimp-1, which suggests that this subset of GC B cells might be gained (A. Shaffer et al., unpublished data). Changes in mRNA were committed to a plasma cell fate. observed that are consistent with a Part of this Blimp-1+ population plasma cell phenotype, including continues to proliferate, suggesting antibody secretion (increases in Ig, that mitogenic signals within the J chain and heat-shock protein 70), GC override Blimp-1–dependent cessation of cell cycle and in- repression of c-Myc and other pro- creased apoptosis (decreases in c- liferation genes at this stage11. Myc, c-Myc targets and the anti- Little is known about signals that apoptotic protein A1; increases in induce Blimp-1, although signal the cyclin-dependent kinase inhibi- transducer and activator of tran- tors p18 and p21 and the apoptotic scription 3 (STAT3) may be proteins GAD153 and GADD45). involved, either directly or indi- We also observed that Blimp-1 rectly, in activating its expres- represses gene expression pro- sion59. Microarray analyses grams required for activated B cells showed that the gene that encodes and/or GC B cells. Expression of Blimp-1 (PRDM1) is repressed by multiple mRNAs encoding pro- BCL-660 and two BCL-6 response teins, including Syk, Btk, Blnk (B elements have been mapped within cell linker protein), Igα, CD19, the PRDM1 gene, confirming that CD20, CD22 and CD45, involved PRDM1is a direct target of BCL-6 in signaling from the BCR was repression (C. Tunyaplin, and K. decreased. Blimp-1 also down-reg- © http://immunol.nature.com Group 2001 Nature Publishing Calame, unpublished data). Re- ulated BCL-6, a protein which is pression of Prdm1 by BCL-6 was required for GC B cells69,70, and also confirmed in studies that other genes required for T cell–B showed ectopic expression of cells interactions and/or GC func- BCL-6 inhibited expression of tions, such as isotype-switch Blimp-1 and differentiation of pri- recombination and somatic hyper- mary B cells to antibody-secreting mutation. These genes included cells59. Thus, one function of BCL- those that encode CD86, AID71, 6 in GC B cells is to repress DNAPKcs, Ku70, Ku86, BSAP, Prdm1, thus inhibiting Blimp- CIITA and STAT6. Thus, Blimp-1 1–dependent terminal differentia- determines plasma cell fate by trig- tion. Signaling from CD40, which gering two programs: inhibition of inhibits plasma cell differentiation, earlier or GC functions and activa- also causes down-regulation of Figure 3. Regulatory cascades in plasma cell development.Transcription tion of plasma cell functions. The 61 factors required for earlier B cells are shown in yellow; those required for plas- Blimp-1 mRNA , emphasizing the ma cells are shown in blue.Targets that are activated by a particular factor are reciprocal ability of Blimp-1 to strong association between Blimp- indicated by arrows; targets that are repressed are indicated by bars.The X indi- repress BCL6 creates a feedback 1 and plasmacytic differentiation. cates degradation of BCL-6 protein after BCR signaling in the GC. loop that enforces strict control Three gene targets for Blimp- over the B cell fate decision to 1–dependent repression have been become a plasma cell. As long as identified; they explain important aspects of the plasma cell pheno- BCL-6 is expressed in a GC B cell, Blimp-1 expression and plasma- type induced by Blimp-1. Blimp-1 represses c-Myc transcription62, cytic differentiation is blocked; if Prdm1 becomes activated, howev- which provides an explanation for cessation of the cell cycle in plas- er, BCL6 will be repressed, which leads to irreversible plasmacytic ma cells, as c-Myc is required for cell proliferation and growth. differentiation. MHC2TA promoter III (the promoter responsible for regulating Mice that lack IRF4 cannot mount an antibody response and lack CIITA in B cells) is also directly repressed by Blimp-155. Repression plasma cells and cytotoxic T cells72, which suggests an important role of CIITA by Blimp-1 explains the down-regulation of MHC class II for IRF4 in plasma cell development and/or function. This IRF family in plasma cells. Finally, Blimp-1 represses Pax5, which encodes protein was identified as a partner for the Ets family protein PU.112 and BSAP (K.-I. Lin et al., unpublished data). BSAP is required for com- by its homology to other IRF proteins73. IRF4 can also associate with mitment to the B cell lineage, for B cell development in the bone mar- the Ets family protein SpiB, the helix-loop-helix protein E47, STAT674

and zinc-finger proteins BCL-6 and Blimp-175,76. Known targets acti- cells, represses XBP150 and lack of XBP-1 also blocks plasma cell vated by complexes that contain IRF4 include κ and λ light chain development14. enhancers and the human CD23 promoter. BCL-6– and/or BSAP-dependent repression is probably relieved Regulation of IRF4 appears to be complex. It is expressed primar- by some appropriate combination of signals that occurs when the GC ily in lymphoid cells and in B cells IRF4 expression depends on Rel reaction is complete. This signal may involve BCR-dependent activa- proteins77. Also, in B cells its expression pattern parallels that of tion via surface Ig with a high affinity for antigen. Signals from BCR Blimp-1: it is increased in a plasma cell–like subset of GC cells and are known to cause mitogen-activated protein kinase–dependent in plasma cells13. However, IRF4 activity is also post-transcriptional- phosphorylation and subsequent degradation of BCL-685. This would ly regulated in several ways. For some sites, binding to DNA is be expected to cause derepression of Blimp-1, thus initiating plasma enhanced when IRF4 is associated with another protein, such as cell development. Appropriate combinations of cytokines are proba- PU.178. Its transcriptional activity is negatively regulated by BCL-6 bly also important. IL-6 induces XBP-152 and other cytokines and and Blimp-175,76. TNFR signals are also likely to play a role in inducing XBP-1, IRF4 The predominant NFAT proteins in lymphocytes are NFATc1 and and/or Blimp-1. Blimp-1 would then repress the genes that encode NFATc279. Mice with lymphocytes that are doubly deficient in BCL-6, AID, CIITA, BSAP and other genes required for GC func- NFATc1 and NFATc2 have T cells with multiply defective effector tions, thus inhibiting GC reactions and ensuring irreversible plasma functions, B cells that are hyperactivated and increased serum IgG1 cell development. Repression of PAX5 by Blimp-1 (K.-I. Lin, C. and IgE as well as plasma cell expansion80. For plasma cells, these Angelin-Duclos and K. Calame, unpublished data) is probably criti- results suggest either a B cell–intrinsic mechanism by which NFATc1 cal to drive plasma cell development because it relieves BSAP-depen- and NFATc2 inhibit plasma cell development, or a mechanism by dent repression of XBP150, which is required for plasma cell func- which T cells that lack NFATc1 and NFATc2 cannot appropriately tion14. It also relieves BSAP-dependent repression of IgH and IgJ regulate plasma cell development. Given the current understanding chain64 transcription, which leads to IgM secretion. Repression of c- that BCL-6 keeps GC B cells from premature plasma cell develop- Myc by Blimp-1 is important for cessation of cell cycle67 in plasma ment by repressing Prdm1, it will be interesting to determine whether cells and repression of other targets, including Bcl-2 family member BCL-6 expression is normal in B cells from mice that lack NFATc1 A1 and MHC2TA, also contributes to the plasma cell phenotype. and NFATc2 expression by their lymphocytes. Many interesting issues remain unanswered at this point: some may Octamer sites are important for transcriptional activity of heavy yield answers rapidly, whereas others may be more challenging. It and light chain promoters and for the 3′ Cα enhancer. Octamer-bind- should be relatively easy to determine whether certain features of this ing factor 1 (Oct-1), which is ubiquitously expressed, and Oct-2, model are correct and to expand some aspects of it. For example, do which has more lymphoid-restricted expression, bind these sites and XBP-1, Blimp-1 and IRF4 function in a single pathway or parallel both associate with B cell–specific octamer cofactor (OCA-B)81. Mice pathways? What causes induction of IRF4, Blimp-1 and XBP-1 in with lymphocytes that lack Oct-2 have normal B cell development, MZ B cells after antigen stimulation? Does BCL-6 repress IRF4 or © http://immunol.nature.com Group 2001 Nature Publishing but the B cells do not respond to polyclonal mitogens and undergo G1 XBP1 as well as Prdm1 in GC B cells? What are the relevant targets arrest rather than proliferation, which presumably accounts for the of XBP-1 and IRF4 in plasmacytic differentiation? It will also be reduced serum Ig in these mice82. Mice that lack OCA-B do not make interesting, but more challenging, to identify individual signals and GCs, but plasma cell development is not affected83. Although B cells particular signal strengths or combinations of signals that determine that lack Oct-2 or OCA-B show defective developmental stages developmental fate decisions. These should include developmental before plasma cell differentiation, the role of octamer proteins in Ig fate choices between B1 and B2 cells, MZ and follicular B cells, plas- promoters and the 3′ Cα enhancer argue that that Oct-1 or Oct-2, ma cells and memory cells after the GC reaction and plasma cells and along with OCA-B, may be important in plasma cells. The occurrence maintenance of memory cells upon secondary stimulation of memory of plasma cells in the absence of OCA-B is not consistent with this cells by antigen. idea, but there have been suggestions of additional, currently uniden- 84 tified, octamer cofactors that may be important in plasma cells. Acknowledgments I thank members of my laboratory and many other colleagues for helpful discussions. Unifying concepts Supported by grant numbers GM29361 and AI43576. Because several key transcriptional regulators involved in plasma cell 1. Sze, D. M.,Toellner, K. M., Garcia de Vinuesa, C.,Taylor, D. R. & MacLennan, I. C. Intrinsic constraint development have been identified, it is important to speculate on how on plasmablast growth and extrinsic limits of plasma cell survival. J. Exp. Med. 192, 813–821 (2000). their functions may be related. Although there is no direct evidence, it 2. Chen-Bettecken, U.,Wecker, E. & Schimpl,A.Transcriptional control of µ- and κ-gene expression in resting and bacterial lipopolysaccharide-activated normal B cells. Immunobiology 174, 162–176 (1987). seems logical that a single developmental sequence may exist in which 3. Takagaki,Y. & Manley, J. L. Levels of polyadenylation factor CstF-64 control IgM heavy chain mRNA these regulators cooperate to bring about plasmacytic differentiation. accumulation and other events associated with B cell differentiation. Mol. Cell. 2, 761–771 (1998). 4. Silacci, P., Mottet,A., Steimle,V., Reith,W. & Mach, B. Developmental extinction of major histocom- We suggest the following working model to integrate the regulation patibility complex class II gene expression in plasmocytes is mediated by silencing of the transac- and action of three major transcription factors: XBP-1, Blimp-1 and tivator gene CIITA. J. Exp. Med. 180, 1329–1336 (1994). 5. Sanderson, R. D., Lalor, P. & Bernfield, M. B lymphocytes express and lose syndecan at specific IRF4 (Fig. 3). stages of differentiation. Cell. Reg. 1, 27–35 (1989). Because these regulators appear to be required for TI and TD 6. Hargreaves, D. C. et al. A coordinated change in chemokine responsiveness guides plasma cell movements. J. Exp. Med. 194, 45–56 (2001). responses, appropriate signals by antigen via BCR probably induce 7. Nutt, S. L., Heavey, B., Rolink,A. G. & Busslinger, M. Commitment to the B-lymphoid lineage XBP-1, Blimp-1 and IRF4 in naïve MZ or follicular B cells. This depends on the transcription factor Pax5. Nature 401, 556–562 (1999). 8. Nutt, S. L., Eberhard, D., Horcher, M., Rolink,A. G. & Busslinger, M. Pax5 determines the identity of would allow MZ B cells to develop rapidly into plasma cells. B cells from the beginning to the end of B-lymphopoiesis. Int. Rev. Immunol. 20, 65–82 (2001). However, terminal differentiation is inhibited as follicular B cells 9. Barberis,A.,Widenhorn, K.,Vitelli, L. & Busslinger, M.A novel B-cell lineage-specific transcription factor present at early but not late stages of differentiation. Genes Dev. 4, 849–859 (1990). enter the GC. This is due, at least in part, to repression of Blimp-1 by 10. Turner, C.A., Jr., Mack, D. H. & Davis, M. M. Blimp-1, a novel zinc finger-containing protein that can BCL-659,60. It may be that XBP1 and/or IRF4 are also repressed by drive the maturation of B lymphocytes into immunoglobulin-secreting cells. Cell 77, 297–306 (1994). 11. Angelin-Duclos, C., Cattoretti, G., Lin, K. I. & Calame, K. Commitment of B lymphocytes to a plas- BCL-6. In addition, BSAP, which is expressed and required in GC B ma cell fate is associated with blimp-1 expression in vivo. J. Immunol. 165, 5462–5471 (2000).

12. Eisenbeis, C. F., Singh, H. & Storb, U. Pip, a novel IRF family member, is a lymphoid-specific, PU. 1- 50. Reimold,A. M. et al. Transcription factor B cell lineage-specific activator protein regulates the gene dependent transcriptional activator. Genes Dev. 9, 1377–1387 (1995). for human X-box binding protein 1. J. Exp. Med. 183, 393–401 (1996). 13. Falini, B. et al. A monoclonal antibody (MUM1p) detects expression of the MUM1/IRF4 protein in 51. Wallin, J. J., Gackstetter, E. R. & Koshland, M. E. Dependence of BSAP repressor and activator func- a subset of germinal center B cells, plasma cells, and activated T cells. Blood 95, 2084–2092 (2000). tions on BSAP concentration. Science 279, 1961–1964 (1998). 14. Reimold,A. M. et al. Plasma cell differentiation requires the transcription factor XBP-1. Nature 52. Wen, X. et al. Identification of c-myc promoter-binding protein and X-box binding protein 1 as 412, 300–307 (2001). interleukin-6 target genes in human multiple myeloma cells. Int. J. Oncol. 15, 173–178 (1999). 15. Macpherson,A. J. et al. IgA production without µ or δ chain expression in developing B cells. 53. Yoshida, H. et al. ATF6 activated by proteolysis binds in the presence of NF-Y (CBF) directly to Nature Immunol. 2, 625–631 (2001). the cis-acting element responsible for the mammalian unfolded protein response. Mol. Cell. Biol. 20, 16. Fu,Y. X. & Chaplin, D. D. Development and maturation of secondary lymphoid tissues. Annu. Rev. 6755–6767 (2000). Immunol. 17, 399–433 (1999). 54. Schliephake, D. E. & Schimpl,A. Blimp-1 overcomes the block in IgM secretion in lipopolysaccha- 17. Loder, F. et al. B cell development in the spleen takes place in discrete steps and is determined by ride/anti- µ F(ab′)2-co-stimulated B lymphocytes. Eur. J. Immunol. 26, 268–271 (1996). the quality of B cell receptor-derived signals. J. Exp. Med. 190, 75–89 (1999). 55. Piskurich, J. F. et al. BLIMP-I mediates extinction of major histocompatibility class II transactivator 18. Martin, F. & Kearney, J. F. B-cell subsets and the mature preimmune repertoire. Marginal zone and expression in plasma cells. Nature Immunol. 1, 526–532 (2000). B1 B cells as part of a “natural immune memory”. Immunol. Rev. 175, 70–79 (2000). 56. Keller,A. D. & Maniatis,T. Only two of the five zinc fingers of the eukaryotic transcriptional repres- 19. Oliver,A. M., Martin, F. & Kearney, J. F. IgMhighCD21high lymphocytes enriched in the splenic marginal sor PRDI-BF1 are required for sequence-specific DNA binding. Mol. Cell. Biol. 12, 1940–1949 (1992). zone generate effector cells more rapidly than the bulk of follicular B cells. J. Immunol. 162, 57. Ren, B., Chee, K. J., Kim,T. H. & Maniatis,T. PRDI-BF1/Blimp-1 repression is mediated by corepres- 7198–7207 (1999). sors of the Groucho family of proteins. Genes Dev. 13, 125–137 (1999). 20. Macpherson,A. J. et al. A primitive T cell-independent mechanism of intestinal mucosal IgA 58. Yu, J.,Angelin-Duclos, C., Greenwood, J., Liao, J. & Calame, K.Transcriptional repression by blimp-1 responses to commensal bacteria. Science 288, 2222–2226 (2000). (PRDI-BF1) involves recruitment of histone. Mol. Cell. Biol. 20, 2592–2603 (2000). 21. Fagarasan, S.,Watanabe, N. & Honjo,T. Generation, expansion, migration and activation of mouse 59. Reljic, R.,Wagner, S. D., Peakman, L. J. & Fearon, D.T. Suppression of signal transducer and activator B1 cells. Immunol. Rev. 176, 205–215 (2000). of transcription 3-dependent B lymphocyte terminal differentiation by BCL-6. J. Exp. Med. 192, 22. Dal Porto, J. M., Haberman,A. M., Shlomchik, M. J. & Kelsoe, G.Antigen drives very low affinity B 1841–1848 (2000). cells to become plasmacytes and enter germinal centers. J. Immunol. 161, 5373–5381 (1998). 60. Shaffer,A. L. et al. BCL-6 represses genes that function in lymphocyte differentiation, inflammation, 23. Tomayko, M. M. & Cancro, M. P. Long-lived B cells are distinguished by elevated expression of A1. and cell cycle control. Immunity 13, 199–212 (2000). J. Immunol. 160, 107–111 (1998). 61. Randall,T. D. et al. Arrest of B lymphocyte terminal differentiation by CD40 signaling: mechanism 24. Lam, K. P., Kuhn, R. & Rajewsky, K. In vivo ablation of surface immunoglobulin on mature B cells by for lack of antibody-secreting cells in germinal centers. Immunity 8, 733–742 (1998). inducible gene targeting results in rapid cell death. Cell 90, 1073–1083 (1997). 62. Lin,Y.,Wong, K. & Calame, K. Repression of c-myc transcription by Blimp-1, an inducer of terminal 25. Cariappa,A. et al. The follicular versus marginal zone B lymphocyte cell fate decision is regulated B cell differentiation. Science 276, 596–599 (1997). by Aiolos, Btk, and CD21. Immunity 14, 603–615 (2001). 63. Usui,T. et al. Overexpression of B cell-specific activator protein (BSAP/Pax-5) in a late B cell is 26. Wang, J. H. et al. Aiolos regulates B cell activation and maturation to effector state. Immunity 9, sufficient to suppress differentiation to an Ig high producer cell with plasma cell phenotype. J. 543–553 (1998). Immunol. 158, 3197–3204 (1997). 27. Khan,W. N. et al. Defective B cell development and function in Btk-deficient mice. Immunity 3, 64. Rinkenberger, J. L.,Wallin, J. J., Johnson, K.W. & Koshland, M. E.An interleukin-2 signal relieves BSAP 283–299 (1995). (Pax5)-mediated repression of the immunoglobulin J chain gene. Immunity 5, 377–386 (1996). 28. Thompson, J. S. et al. Baff-R, a newly identified TNF receptor that specifically interacts with BAFF. 65. Singh, M. & Birshtein, B. K. NF-HB (BSAP) is a repressor of the murine immunoglobulin heavy- Science 293, 2108–2111 (2001). chain 3′αenhancer at early stages of B-cell differentiation. Mol. Cell. Biol. 13, 3611–3622 (1993). 29. Marsters, S.A. et al. Interaction of the TNF homologues BLyS and APRIL with the TNF receptor 66. Waldburger, J. M. et al. Lessons from the bare lymphocyte syndrome: molecular mechanisms regu- homologues BCMA and TACI. Curr. Biol. 10, 785–788 (2000). lating MHC class II expression. Immunol. Rev. 178, 148–165 (2000). 30. Do, R. & Chen-Kiang, S. Mechanism of BLyS action in B cell immunity. Cytok. Growth Factor Rev. 67. Lin, K. I., Lin,Y. & Calame, K. Repression of c-myc is necessary but not sufficient for terminal dif- 199, 1–8 (2001). ferentiation of B lymphocytes in vitro. Mol. Cell. Biol. 20, 8684–8695 (2000). 31. McHeyzer-Williams, L. J., Driver, D. J. & McHeyzer-Williams, M. G. Germinal center reation. Curr. 68. Horcher, M., Souabni,A. & Busslinger, M. Pax5/bsap maintains the identity of B cells in late b lym- Opin. Hematol. 8, 52–59 (2001). phopoiesis. Immunity 14, 779–790 (2001). 32. Merville, P. et. al. Bcl-2+ tonsillar plaoma cells are rescued from apoptosis by bone marrow fibrob- 69. Dent,A. L., Shaffer,A. L.,Yu, X.,Allman, D. & Staudt, L. M. Control of inflammation, cytokine expres- lasts. J. Exp. Med. 183, 227–236 (1996). sion, and germinal center formation by BCL-6. Science 276, 589–592 (1997). 33. Slifka, M. K.,Antia, R.,Whitmire, J. K. & Ahmed, R. Humoral immunity due to long-lived plasma 70. Ye, B. H. et al. The BCL-6 proto-oncogene controls germinal-centre formation and Th2- type cells. Immunity 8, 363–372 (1998). inflammation. Nature Genet. 16, 161–170 (1997). 34. Smith, K. G., Light,A., Nossal, G. J. & Tarlinton, D. M.The extent of affinity maturation differs 71. Muramatsu, M. et al. Class switch recombination and hypermutation require activation-induced between the memory and antibody-forming cell compartments in the primary immune response. cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102, 553–563 (2000). © http://immunol.nature.com Group 2001 Nature Publishing EMBO J. 16, 2996–3006 (1997). 72. Mittrucker, H.W. et al. Requirement for the transcription factor LSIRF/IRF4 for mature B and T 35. Arpin, C. et al. Generation of memory B cells and plasma cells in vitro. Science 268, 720–722 (1995). lymphocyte function. Science 275, 540–543 (1997). 36. Stuber, E. & Strober,W.The T cell-B cell interaction via OX40-OX40L is necessary for the T 73. Matsuyama,T. et al. Molecular cloning of LSIRF, a lymphoid-specific member of the interferon regu- cell–dependent humoral immune response. J. Exp. Med. 183, 979–989 (1996). latory factor family that binds the interferon-stimulated response element (ISRE). Nucleic Acids 37. Choe, J. & Choi,Y. S. IL-10 interrupts memory B cell expansion in the germinal center by inducing Res. 23, 2127–2136 (1995). differentiation into plasma cells. Eur. J. Immunol. 28, 508–515 (1998). 74. Mamane,Y. et al. Interferon regulatory factors: the next generation. Gene 237, 1–14 (1999). 38. Zhang, X. et al. The distinct roles of T cell-derived cytokines and a novel follicular dendritic cell- 75. Gupta, S., Jiang, M.,Anthony,A. & Pernis,A. B. Lineage-specific modulation of interleukin 4 signaling signaling molecule 8d6 in germinal center-B cell differentiation. J. Immunol. 167, 49–56 (2001). by interferon regulatory factor 4. J. Exp. Med. 190, 1837–1848 (1999). 39. Kawano, M. M., Mihara, K., Huang, N.,Tsujimoto,T. & Kuramoto,A. Differentiation of early plasma 76. Gupta, S.,Anthony,A. & Pernis,A. B. Stage-specific modulation of IFN-regulatory factor 4 function cells on bone marrow stromal cells requires interleukin-6 for escaping from apoptosis. Blood 85, by kruppel-type zinc finger proteins. J. Immunol. 166, 6104–6111 (2001). 487–494 (1995). 77. Grumont, R. J. & Gerondakis, S. Rel induces interferon regulatory factor 4 (IRF-4) expression in 40. Morse, L., Chen, D., Franklin, D., Xiong,Y. & Chen-Kiang, S. Induction of cell cycle arrest and B cell lymphocytes. Modulation of interferon-regulated gene expression by rel/nuclear factor κB. J. Exp. terminal differentiation by CDK inhibitor p18(INK4c) and IL-6. Immunity 6, 47–56 (1997). Med. 191, 1281–1292 (2000). 41. Gado, K., Domjan, G., Hegyesi, H. & Falus,A. Role of interleukin-6 in the pathogenesis of multiple 78. Brass,A. L., Kehrli, E., Eisenbeis, C. F., Storb, U. & Singh, H. Pip, a lymphoid-restricted IRF, contains a myeloma. Cell. Biol. Int. 24, 195–209 (2000). regulatory domain that is important for autoinhibition and ternary complex formation with the 42. Garcia De Vinuesa, C. et al. Dendritic cells associated with plasmablast survival. Eur. J. Immunol. 29, Ets factor PU. 1. Genes Dev. 10, 2335–2347 (1996). 3712–3721 (1999). 79. Rao,A., Luo, C. & Hogan, P. G.Transcription factors of the NFAT family: regulation and function. 43. McHeyzer-Williams, M. G. & Ahmed, R. B cell memory and the long-lived plasma cell. Curr. Opin. Annu. Rev. Immunol. 15, 707–747 (1997). Immunol. 11, 172–179 (1999). 80. Peng, S. L., Gerth,A. J., Ranger,A. M. & Glimcher, L. H. NFATc1 and NFATc2 together control both 44. Maruyama, M., Lam, K. P. & Rajewsky, K. Memory B-cell persistence is independent of persisting T and B cell activation and differentiation. Immunity 14, 13–20 (2001). immunizing antigen. Nature 407, 636–642 (2000). 81. Matthias, P. Lymphoid-specific transcription mediated by the conserved octamer site: who is doing 45. Driver, D., McHeyzer-Williuams, L., C., M., Stetson, D. & McHeyzer-Williams, M. Development and what? Semin. Immunol. 10, 155–163 (1998). maintenance of a B220-memory B cell compartment. J. Immunol. 167, 1393–1405 (2001). 82. Corcoran, L. M. & Karvelas, M. Oct-2 is required early in T cell-independent B cell activation for 46. Arpin, C., Banchereau, J. & Liu,Y. J. Memory B cells are biased towards terminal differentiation: a G1 progression and for proliferation. Immunity 1, 635–645 (1994). strategy that may prevent repertoire freezing. J. Exp. Med. 186, 931–940 (1997). 83. Qin, X. F., Reichlin,A., Luo,Y., Roeder, R. G. & Nussenzweig, M. C. OCA-B integrates B cell antigen 47. Silvy,A., Lagresle, C., Bella, C. & Defrance,T.The differentiation of human memory B cells into spe- receptor-, CD40L- and IL 4-mediated signals for the germinal center pathway of B cell develop- cific antibody- secreting cells is CD40 independent. Eur. J. Immunol. 26, 517–524 (1996). ment. EMBO J. 17, 5066–5075 (1998). 48. Liou, H. C. et al. A new member of the leucine zipper class of proteins that binds to the HLA DR 84. Schubart, K. et al. B cell development and immunoglobulin gene transcription in the absence of α promoter. Science 247, 1581–1584 (1990). Oct-2 and OBF-1. Nature Immunol. 2, 69–74 (2001). 49. Reimold,A. M. et al. An essential role in liver development for transcription factor XBP-1. Genes 85. Niu, H.,Ye, B. H. & Dalla-Favera, R.Antigen receptor signaling induces MAP kinase-mediated phos- Dev. 14, 152–157 (2000). phorylation and degradation of the BCL-6 transcription factor. Genes Dev. 12, 1953–1961 (1998).

1108 nature immunology • volume 2 no 12 • december 2001 • http://immunol.nature.com